Discussion 275 



in clinical practice. I believe this is, to say the least, debatable. As Dr. 

 Barber has pointed out, this may be true for streptomycin and tubercu- 

 losis; it was probably true for sulphonamide and streptococcal or gono- 

 coccal infections. For the rest, however, originally sensitive bacterial 

 species are not becoming resistant to antibiotics in consequence of 

 chemotherapy. After ten years of intensive use, I don't know of a single 

 documented instance of a penicillin-resistant group A streptococcus, 

 pneumococcus, gonococcus or treponeme appearing as the result of 

 treatment. The clinical problem is a very different one. It is the problem 

 of treatment failure, rather than bacterial resistance : treatment failure 

 due to a number of factors, and often unexplained, but not involving 

 drug resistance in the sense that we have been discussing. Even the 

 problem of penicillin-resistant staphylococcal infection, as Dr. Barber 

 has pointed out, does not involve the development of resistance in a 

 strain which was originally sensitive, but, rather, super-infection with a 

 resistant strain prevalent in hospitals. This poses manj^ problems with 

 respect to the genesis and ecology of these penicillin-resistant staphy- 

 lococci, and their survival value in competition with susceptible strains. 

 These problems are, however, quite different from those under considera- 

 tion these past few days. 



Cavalli-Sforza: Visconti and Szybalski made experiments similar to 

 those you have reported, Dr. Barber. Their point was similar to yours 

 in that you cannot hope to recover penicillinase-producing mutants, 

 except at the very borderline between the highest tolerated concentration 

 and the lowest lethal one. Those workers seemed to get things which 

 were producing a very decent amount of penicillinase. 



Barber: In one experiment, Szybalski did get variants producing a fair 

 amount of penicillinase, but he did no studies of phage typing of parent 

 and child at the end. He got them on gradient plates, but did not 

 publish details, w ith the exception that he said that these strains were 

 producing large amounts of penicillinase. The phage tj^e of these very 

 rarely isolated colonies on gradient plates and the parent strains that 

 they came from were not checked. 



Cavalli-Sforza : I think it may be worth emphasizing that the effecti\'e- 

 ness of selection for penicillinase-producers only at borderline concen- 

 trations has some relevance for the clinical aspects, because on super- 

 infection you cannot usually expect large masses of superinfecting 

 bacteria, but only few, a fact which makes the penicillin concentration 

 even more critical, and therefore you are likely to get superinfection by 

 the penicillinase-producing strains only when very small amounts of 

 penicillin are in circulation. 



Hayes: I would like to suggest a possible genetic model for what is 

 happening in your system. Dr. Barber: you may have two mutations 

 involved ; mutation 1 would be a mutation to the constitutive production 

 of small amounts of penicillinase; mutation 2 would determine the 

 possibility of induction on the Pollock model in B. subtilis. Mutation 2 

 without mutation 1 is useless, because there is no base-line to work on. 

 When you try to select for penicillinase-producing strains, there may be, 

 in the population, a very small minority of cells which can constitutively 



